The cutting stock problem in the canvas industry

Abstract

A number of variations of the cutting stock problem arise in the canvas industry. Due to the nature of operation in the industry the majority of these problems cannot be solved by the conventional cutting stock solution procedures. The problems are two-dimensional in form but involve minimizing the utilized length of a single stock-roll rather than the conventional objective of minimizing waste over multiple stockplates. Small order quantities are also involved, which cause difficulties in the conventional methodologies. Several heuristics to produce a pattern for all components of a unit, such as a tent, are given. These heuristics produce solutions requiring about the same amount of canvas as the current best manual solutions. However, the heuristics' solutions have proven to have significant advantages over the manual solutions in both the time to generate the solution and the type of pattern produced. The heuristics involve a modified form of the conventional three-stages Gilmore-Gomory algorithm combined with integer programming.